Re: Comments on the complexity of the JSON-LD spec

On 6/23/11 6:15 AM, Brian Peterson wrote:
> What is the point of trying to encode LD into JSON? The fact that JSON fits
> well into JS? I guess that's ok if your target is just JS, but JSON is used
> in a number of other languages as well.
>
> So a JSON-LD specification that is straight-forward and simple will make it
> a lot easier for services to use RDF-based LD. They don't even need to want
> to use RDF, just LD and JSON, but the RDF will still be there.
Brian,

A really important point re. the excerpt above (BTW - I agree with the 
essence of your well articulated post). You can have a JSON based graph 
where each Object is endowed with URI based Name that resolves to its 
negotiated representation(s). Said representations take the form of an 
EAV/SPO graph pictorial. Syntax for constructing representations are 
naturally varied since there a many syntaxes for representing graphs.

The statement above isn't about RDF, that's Linked Data.

Linked Data simply needs URI based Identifiers to resolve to 
representations of their Referents.

RDF is one of many routes to creating Linked Data with varying levels of 
semantic fidelity.


As you've articulated, ditto Glenn, and others. JSON-LD, I hope, is 
supposed to be about Linked Data not RDF. Any RDFer worth his or her 
salt can make higher semantic fidelity based Linked Data from a basic 
Linked Data graph constructed using a JSON based syntax.

Manu: what's the goal of JSON-LD? Is it to mirror RDF or deliver a 
beachhead for a broader base of developers  who don't give a darn about 
RDF, at least right now?

-- 

Regards,

Kingsley Idehen	
President&  CEO
OpenLink Software
Web: http://www.openlinksw.com
Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca: kidehen

Received on Thursday, 23 June 2011 08:00:34 UTC